Psychometric Properties
Running Head:
1
Psychometric Properties of the MEIM
Evaluating the Psychometric Properties of the Multigroup Ethnic
Identity Measure (MEIM) within the United Kingdom
Stanley O. Gaines, Jr., David Bunce, Toby Robertson, and Barlow
Wright
with
Yoeri Goossens, Daljeet Heer,
Amarpreet Lidder, Amardeep Mann, and Sonia Minhas
Brunel University
Author Posting. (c) Taylor & Francis Group, LLC, 2010.
This is the author's version of the work. It is posted here by
permission of Taylor & Francis Group, LLC, for personal use,
not for redistribution. The definitive version was published
in Identity, Volume 10 Issue 1, January 2010.
doi:10.1080/15283481003676176(http://dx.doi.org/10.1080/
15283481003676176). The authors thank Seth Schwartz and three
anonymous reviewers for their constructive comments on earlier
versions of this paper. Correspondence should be addressed to
Stanley O. Gaines, Jr., Department of Psychology, School of
Social Sciences, Brunel University, Uxbridge, Middlesex UB8
3PH. United Kingdom. e-mail: [email protected]
Psychometric Properties
2
Abstract
In the present study, we examined the psychometric properties of
the Multigroup Ethnic Identity Measure (MEIM; Phinney, 1992;
Phinney & Alipuria, 1990) among an ethnically diverse sample
within the United Kingdom.
In initial analyses, we evaluated
the goodness-of-fit of a 1-factor model (i.e., global ethnic
identity) and the goodness-of-fit of a 2-factor model (i.e.,
correlated but distinct Exploration and Commitment components).
Results of initial confirmatory factor analyses led us to reject
both the 1-factor and 2-factor models.
Results of subsequent
exploratory and confirmatory factor analyses revealed a 3-factor
structure (i.e., correlated but distinct Behavioral, Cognitive,
and Affective components of ethnic identity) among the sample as
a whole (n = 234) and among Asian Indian persons (n = 88) in
particular, though results were mixed among White U.K./Irish
persons (n = 54).
Implications for the study of ethnicity-
related concepts in the increasingly multi-cultural U.K. are
discussed.
KEYWORDS:
Ethnic identity; ego psychology; identity status;
MEIM; United Kingdom.
Psychometric Properties
3
Evaluating the Psychometric Properties of the Multigroup Ethnic
Identity Measure (MEIM) within the United Kingdom
Within the United States, the 1990s were hailed as the
“decade of ethnicity” in psychology (Shweder & Sullivan, 1993,
p. 517).
During the 1990s, American psychologists increasingly
distinguished among ethnicity-related constructs such as
minority versus majority status, culture, and identity (Phinney,
1996).
In particular, the construct of ethnic identity (i.e.,
individuals‟ self-categorization in, and psychological
attachment toward, the ethnic groups to which they belong;
Phinney, 1990) has received considerable attention among
American psychologists since the early 1990s (Verkuyten, 2005).
In contrast, the study of ethnic identity in the United
Kingdom is in its relative infancy.
Bhui et al. (2005)
ostensibly examined cultural identity among persons of African,
Asian (specifically Bangladeshi), and European (specifically
British) descent; but they actually examined acculturation,
rather than ethnic identity, as a predictor of health status.
Berry, Phinney, Sam, and Vedder (2006) clearly examined ethnic
identity as part of a battery of constructs in a 17-nation study
that included the U.K.; but their U.K. sample was limited to
persons of Asian (specifically Indian) descent.
We do not know
of any published study in which ethnic identity, as distinct
from acculturation, has been measured across multiple ethnic
groups within the U.K.
As the U.K. increasingly has become
Psychometric Properties
4
ethnically diverse, social scientists and students in the U.K.
increasingly have begun to grapple with ethnicity-related issues
such as ethnic identity (Alexander, 2006).
Empirical research
on ethnic identity is needed, not just within the higher
education sector, but throughout U.K. society as a whole.
The Multigroup Ethnic Identity Measure (MEIM; Phinney,
1992; Phinney & Alipuria, 1990) has emerged as the most widely
used measure of ethnic identity within as well as outside the
U.S. (for examples within the U.S., see Avery, Tonidandel,
Thomas, Johnson, & Mack, 2007; Gaines, Marelich, Bledsoe,
Steers, Henderson, Granrose, et al., 1997; Ponterotto, Gretchen,
Utsey, Stracuzzi, & Saya, 2003; for examples outside the U.S.,
see Dandy et al., 2008; Verkuyten & Yildiz, 2006).
In the
present study, we evaluated the psychometric properties of the
12-item version of the MEIM (Roberts et al., 1999) across
multiple ethnic groups within the U.K.
We sought to determine
whether a 1-factor solution (i.e., a single dimension of ethnic
identity; Phinney & Alipuria, 1990) or a 2-factor solution
(i.e., Exploration and Commitment as two correlated yet distinct
dimensions of ethnic identity; Roberts et al., 1999) provided
optimal fit to the data.
Conceptual Origins of the MEIM: Erikson, Marcia, and Phinney
Erikson’s theory of ego psychology.
Baumeister (1997, p.
682) defined identity as “…the [aggregate of] definitions that
are created for and superimposed on the self. . . .”
According
to Erikson‟s (1950, 1968) theory of ego psychology, the
Psychometric Properties
5
development and maintenance of a stable identity is a major task
of adolescence and adulthood.
Individuals‟ success in striving
to develop and maintain a stable identity is known as identity
achievement; whereas individuals‟ failure in striving to develop
and maintain a stable identity is known as identity confusion in
Erikson‟s ego psychology theory.
Identity is one of three major aspects of individuals‟
ethnicity (along with culture and minority versus majority
status; Phinney, 1996).
Especially relevant to the present
study have been researchers‟ attempts to measure individual
differences in ethnic identity (e.g., Phinney & Alipuria, 1990;
Umaña-Taylor, Yazedijan, & Bamaca-Gomez, 2004).
Erikson‟s
(1950, 1968) ego psychology theory suggests that within a given
ethnic group, individuals vary along a continuum ranging from
ethnic identity achievement at the high end to ethnic identity
confusion at the low end.
Marcia’s model of identity statuses.
Marcia (1966, 1980)
was the first identity theorist to propose a model of identity
statuses on the basis of Erikson‟s (1950, 1968) ego psychology
theory (Schwartz, 2001a, b).
First, Marcia extracted the themes
of Exploration (i.e., “the sorting though of multiple
alternatives”; Schwartz, 2001b, p. 11) and Commitment (i.e.,
“the act of choosing one or more alternatives and following
through with them”; Schwartz, 2001b, p. 11) from Erikson‟s
writings.
Subsequently, Marcia developed and tested a taxonomy
of identity statuses reflecting individual differences in levels
Psychometric Properties
6
of Exploration and Commitment.
Ideally, the combination of high versus low levels of
Exploration and high versus low levels of Commitment should
result in four identity statuses:
(1) Identity achievement
(high levels of Exploration and Commitment); (2) identity
moratorium (high level of Exploration and low level of
Commitment); (3) identity foreclosure (low level of Exploration
and high level of Commitment); and (4) identity diffusion (low
levels of Exploration and Commitment).
Marcia (1966, 1980)
argued that identity achievement is the most advanced stage of
identity development; identity moratorium is the next-to-most
advanced stage of identity development; identity foreclosure is
the next-to-least advanced stage of identity development; and
identity diffusion is the least advanced stage of identity
development.
Marcia and his colleagues (e.g., Marcia, 1966,
1967; Marcia & Friedman, 1970; Orlofsky, Marcia, & Lesser, 1973;
Schenkel & Marcia, 1972; Toder & Marcia, 1973) consistently
obtained support for the identity status model among late
adolescents and young adults.
Phinney’s model of stages of ethnic identity development.
Drawing upon Erikson‟s (1950, 1968) theory of ego psychology and
Marcia‟s (1966, 1980) taxonomy of identity statuses, Phinney
(1990) developed a model of stages of ethnic identity
development across the life span.
Phinney distinguished among
the unexamined ethnic identity stage, reflecting identity
diffusion and identity foreclosure; the ethnic identity search
Psychometric Properties
7
or Exploration stage, reflecting identity crisis and moratorium;
and the achieved ethnic identity or Achievement stage.
In
Phinney‟s model, Achievement represents the highest level of
ethnic identity development; Exploration represents an
intermediate level of ethnic identity development; and
unexamined ethnic identity represents the lowest level of ethnic
identity development.
Various identity theorists have recommended that
researchers focus on the continuous dimensions of Exploration
and Commitment, rather than on the identity statuses (e.g.,
Bosma & Kunnen, 2001; Luyckx, Goossens, Soenens, & Beyers, 2006;
Schwartz, 2007).
Consistent with these recommendations,
Phinney‟s (1992; Phinney & Alipuria, 1990) MEIM does not
classify individuals according to discrete stages of ethnic
identity development.
Rather, in its various incarnations, the
MEIM measures individuals‟ Exploration and Commitment regarding
ethnic identity along a continuum (Phinney & Ong, 2007).
Psychometric Properties of the MEIM:
Ethnic Identity as a
Unidimensional versus Bidimensional Construct
The psychometric properties of the MEIM have been evaluated
in several published studies within the U.S. (e.g., Avery et
al., 2007; Gaines et al., 1997; Pegg & Plybon, 2005; Ponterotto
et al., 2003; Reese, Vera, & Paikoff, 1998; Spencer, Icard,
Harachi, Catalano, & Oxford, 2000; Yancey, Aneshensel, &
Driscoll, 2001).
Debates concerning the psychometric properties
of the MEIM usually have focused on the factor structure of the
Psychometric Properties
8
MEIM, especially regarding the presence of one versus two
underlying dimensions.
The presence of one dimension would lend
support to Erikson‟s (1950, 1968) view that individuals vary
along an axis from low to high identity development; whereas the
presence of two dimensions might lend support to Marcia‟s (1966,
1967) view that individuals vary along the orthogonal axes from
low to high Exploration and from low to high Commitment.
Ethnic identity as a unidimensional construct.
Throughout
most of the 1990s, Phinney and her colleagues (e.g., Phinney,
1992; Phinney & Alipuria, 1990; Phinney, Chavira, & Tate, 1993;
Phinney, Ferguson, & Tate, 1997) generally described the MEIM as
a valid, reliable measure of global ethnic identity.
Various
researchers in the U.S. (e.g., Avery et al., 2007; Gaines et
al., 1997; Ponterotto et al., 2003) similarly have concluded
that the MEIM measures overall (i.e., global) ethnic identity.
Results concerning the unidimensionality of the MEIM are not
consistent with Marcia‟s (1966, 1967) model, which originally
served as the point of departure for Phinney‟s (1990) model.
However, results concerning the unidimensionality of the MEIM
are consistent with Erikson‟s (1950, 1968) earlier writings on
identity development.
Ethnic identity as a bidimensional construct.
More
recently, since the late 1990s, Phinney and her colleagues
(e.g., Phinney & Ong, 2007; Roberts et al., 1999) generally have
described the MEIM as a valid, reliable measure of two related
yet separate components of ethnic identity, namely Exploration
Psychometric Properties
9
(a cognitive and developmental component) and Commitment (an
affective component).
Various researchers in the U.S. (e.g.,
Pegg & Plybon, 2005; Reese, Vera, & Paikoff, 1999; Spencer,
Icard, Harachi, Catalano, & Oxford, 2000; Yancey, Aneshensel, &
Driscoll, 2001) and in Australia (e.g., Dandy et al., 2008)
similarly have obtained support for separate Exploration and
Commitment components of the MEIM.
Results concerning the
bidimensionality of the MEIM are consistent with Marcia‟s (1966,
1967) identity status model.
Contradictions regarding the unidimensionality versus
multidimensionality of the MEIM in Phinney’s research.
As we
have noted, since the late 1990s, Phinney‟s research generally
has supported a 2-factor model for the MEIM (Phinney & Ong,
2007; Roberts et al., 1999).
However, Phinney‟s own results in
a recent study across Australia, the U.S., the U.K., and 14
other nations (Berry, Phinney, Sam, & Vedder, 2006) support a
one-factor, rather than a two-factor, structure.
The discrepant
findings across Phinney‟s own studies underscore the need for
overt tests of 1- versus 2-factor models for the MEIM.
Ethnic Group Differences in Factor Structure of the MEIM
Scale(s)
Results of studies by Roberts et al. (1999) and by Avery et
al. (2007), administering the 12-item MEIM to ethnically diverse
samples within the U.S. (i.e., European Americans, African
Americans, Latinas/os, and Asian Americans), indicate that a 2factor structure with Exploration and Commitment as correlated
Psychometric Properties
10
yet distinct components can be applied across ethnic groups.
However, in both studies, the magnitude of factor loadings was
unequal across ethnic groups.
Thus, these studies yielded
configural invariance (i.e., same overall structure) but failed
to yield metric invariance (i.e., equivalent loadings).
Overall, the lack of metric invariance in the factor structure
of the MEIM across ethnic groups suggests that factor
equivalence cannot be taken for granted.
More research appears
to be needed to investigate this issue, especially in contexts
outside the United States.
The issue of equivalence in factor structure across ethnic
groups takes on added importance when one considers that the
only published study in the U.K. using the MEIM (Berry, Phinney,
Sam, & Vedder, 2006) included only one ethnic group (i.e., Asian
Indians) and, thus, did not test for equivalence in factor
structure.
Moreover, the 1-factor structure obtained in that
U.K. sample is at odds with the 2-factor structure obtained by
Pegg and Plybon (2005) using a sample of African American girls,
and with the aforementioned 2-factor structures obtained by
Roberts et al. (1999) and by Avery et al. (2007) using
multiethnic samples in the U.S.
Given such contradictory
findings in previous studies, we made no predictions regarding
the nature of factor structure equivalence of the MEIM across
ethnic groups in the U.K.
Goals of the Present Study
In the present study, we studied a multi-ethnic sample of
Psychometric Properties
students from the U.K.
11
We conducted confirmatory factor
analyses on the sample as a whole, in order to evaluate the
goodness-of-fit of 1-factor (i.e., global ethnic identity) and
2-factor (i.e., Search and Affirmation) structural equation
models using the MEIM. In doing so, we tested the hypothesis
(Phinney & Ong, 2007) that a 2-factor solution would fit the
data significantly better than would a 1-factor solution.
After
deciding how many factors to retain, we conducted multiple-group
confirmatory factor analyses to test the hypothesis (Phinney &
Ong, 2007) that the factor structure of the MEIM would be
invariant across ethnic groups.
Method
Participants
A total of 236 individuals (126 men, 108 women, and 2
individuals who did not report their gender) participated in the
present study.
Participants were recruited within and outside
Brunel University, located in West London.
The mean age of
participants was 23.67 years (SD = 6.68 years).
In terms of
ethnicity, 55.5% of participants were of Asian descent (37.3%
Indian, 9.7% Pakistani, 4.7% Bangladeshi, 1.3% Chinese, and 2.5%
“Asian Other”); 7.2% were of African descent (1.3% Black
Caribbean, 5.9% Black African); 32.2% were of European descent
(22.9% White U.K./Irish, 6.4% White European, and 3.0% “White
Other”); 3.4% were of “Mixed Race” (a generic term covering
various racial heritages); and 1.7% did not report their
ethnicity.1
Psychometric Properties
12
Procedure
Prior to conducting the present study, the first author
obtained ethics approval from a departmental research ethics
subcommittee.
The present study was conducted in accordance
with the Code of Conduct, Ethical Principles and Guidelines of
the British Psychological Society (BPS, 2005).
The BPS ethics
guidelines are similar to APA ethics guidelines in terms of
stringency (see Kimmel, 2004).
Researchers introduced themselves, stated that they were
collecting data for their respective undergraduate theses, and
stated that they were seeking participants for a large-scale
study of personality and personal relationship processes.
Researchers emphasized that participation in the present study
was strictly voluntary (i.e., participants did not receive
money, course credit, or other compensation for taking part in
the study).
All materials were presented in English, and no
translation was needed for any of the participants to understand
the materials.
Participants read and signed informed consent
forms (which explained the purpose of the study in general),
completed survey questionnaires (including demographic items,
measures of ethnic identity, and additional measures of
personality and social behavior that are beyond the scope of the
present paper; Goossens, 2006; Heer, 2006; Lidder, 2006; Mann,
2006; Minhas, 2006), and read debriefing forms (explaining the
purpose of the study in detail).
In general, participants
completed the survey within 30 minutes.
Psychometric Properties
13
Materials
Participants completed the 12-item Multi-Group Ethnic
Identity Measure (MEIM; Roberts et al., 1999).
Each item was
scored according to a 5-point, Likert scale (1 = strongly
disagree, 5 = strongly agree), such that higher scores reflected
greater tendencies for individuals to think of themselves in
terms of their ethnicity and to feel psychologically attached to
the ethnic groups to which they belong.
the following:
Sample items included
“I have spent time trying to find out more about
my ethnic group, such as its history, traditions, and customs”
(Exploration); and “I have a clear sense of my ethnic background
and what it means for me” (Commitment).
Given that a major goal
of the present study was to determine the factor structure of
the MEIM, we report results of factor and reliability analyses
in the Results section.
Results
The goodness of fit of confirmatory and exploratory factor
analyses of the MEIM was assessed via the following indices
where available: (a) Chi-square (χ2), the only index that is
accompanied by a formal statistical test of significance whereby
significant values represent unacceptably fitting models and
nonsignificant models represent acceptably fitting models
(ideally, models whose estimated correlation matrices do not
represent significant departures from actual correlation
matrices yield nonsignificant chi-squares2); (b) chisquare/degrees-of-freedom ratio (χ2/df), a variation on χ2 in
Psychometric Properties
14
which the formal significance test is replaced by a range
whereby a ratio above 3.00 is considered too high, a ratio
approaching but not dropping below 1.00 is considered optimal,
and a ratio below 1.00 is considered “too good to be true” and
thus is too low; (c) root mean square error of approximation
(RMSEA), another variation on χ2 that does not reflect a
significance test but nonetheless can be interpreted such that
values approaching 0 (and, preferably, below .10) are considered
optimal; and comparative fit index (CFI), another variation on χ2
in which the formal significance test is replaced by a range
whereby values approaching 1.00 (and preferably, above .90) are
considered optimal.
Initial Analyses
Confirmatory factor analysis, 1-factor model, all
participants.
The matrix of zero-order correlations among item
scores for the full sample (N = 234) is available from the first
author upon request.
All correlations were positive and
significant (all p‟s < .01).
The correlation matrix was entered
into a confirmatory factor analysis in which we evaluated the
goodness-of-fit of a 1-factor model using LISREL 8.72 (Jöreskog
& Sörbom, 2005a).
In the 1-factor model, none of the
measurement error terms were allowed to correlate; and all
factor loadings were freely estimated.
Results of the
confirmatory factor analysis, using a maximum likelihood
solution, indicated that a 1-factor model generally provided a
Psychometric Properties
15
poor fit to the data (χ2 (54) = 378.39, p < .001; χ2/df = 7.01;
RMSEA = .17; CFI = .93).
Confirmatory factor analysis, 2-factor model, all
participants.
Next, we evaluated the goodness-of-fit of a 2-
factor model.
In the 2-factor model, none of the measurement
error terms were allowed to correlate; the Exploration items
were attached to one factor, the Commitment items were attached
to a second factor; and the two factors were allowed to
correlate.
Results of the confirmatory factor analysis, using a
maximum likelihood solution, indicated that a 2-factor model
generally provided a poor fit to the data χ2 (53) = 296.92, p <
.01; χ2/df = 5.60; RMSEA = .14; CFI = .94).
Comparison of 1-factor and 2-factor models.
A direct
comparison of the 1-factor and 2-factor models indicates that
the 2-factor model provided a significantly better fit than did
the 1-factor model (reduction in χ2 = 81.43; reduction in df = 1;
p < .01).
However, given that neither the 1-factor model nor
the 2-factor model provided an acceptable fit, the results of
this comparison are insufficient to justify accepting either
model (unlike Phinney & Ong, 2007).
In order to obtain a 1-
factor or a 2-factor model with adequate fit to the data, we
would have needed to add several unexpected instances of
uncorrelated measurement error (as did Avery et al., 2007;
Gaines et al., 1997; and Roberts et al., 1999).
Although the
addition of uncorrelated measurement error terms can be
justified in some cases (Kline, 2005), we decided not to employ
Psychometric Properties
this technique.
16
Instead, we conducted an exploratory factor
analysis in order to determine how many factors were measured by
the MEIM (for examples of complementary uses of exploratory and
confirmatory factor analyses, see Brown, 2006).
Subsequent Analyses
Exploratory factor analysis, all participants.
Our
decision to shift from confirmatory to exploratory factor
analysis made it necessary for us to use the predecessor to
LISREL (Jöreskog & Sörbom, 1996a), known as PRELIS (Jöreskog &
Sörbom, 1996b).
PRELIS provides SEM fit indices for the factor
solution extracted. Unlike LISREL (which can be conducted using
a zero-order correlation matrix), PRELIS requires raw data.
Thus, we entered the raw data (available from the first author
upon request) into PRELIS 2.72 (Jöreskog & Sörbom, 2005b).
Results of the exploratory factor analysis, using a maximum
likelihood solution, indicated that no more than six factors
could be extracted from the data (i.e., iterations did not
converge for seven factors).
By default, PRELIS retains the
solution with the highest number of factors that yields a
nonsignificant chi-square.
However, using this default solution
would have led us to retain a 5-factor solution with Heywood
cases (i.e., one or more communality estimates greater that 1.00
and, thus, inadmissible; Thompson, 2004).
We conducted the exploratory factor analysis a second time,
retaining the solution with the highest number of factors
yielding a root mean square error of approximation (RMSEA) lower
Psychometric Properties
17
than .10 (following Phinney & Ong, 2007 see also Kline, 2005) –
in this instance, a 3-factor solution.
The 3-factor solution
generally provided an acceptable fit to the data (χ2 (33) =
81.86, p < .01; χ2/df = 2.48; RMSEA = .08)3.
The resulting
matrix of Promax-rotated factor loadings is shown in Table 1.
Inspection of the factor loadings in Table 1 reveals that using
the criterion of a maximum of 1 item with a factor loading of
.30 or higher per row (Kline, 1994), Items 1, 2, 3, and 4
measure the Behavioral component of ethnic identity; Items 6, 8,
and 9 measure the Cognitive component of ethnic identity; and
Items 7, 10, and 12 measure the Affective component of ethnic
identity (Items 5 and 11 loaded on more than one factor and,
thus, are excluded in interpreting the factors).
Having determined that a three-factor solution was optimal,
we conducted reliability analyses on the resulting subscales.
Results of reliability analyses indicated that the three
subscales were internally consistent, especially given the small
number of items per subscale (Cronbach‟s alphas = .81 for the
Behavioral component, .89 for the Cognitive component, and .89
for the Affective component; see Carmines & Zeller, 1979,
regarding alpha coefficients in reliability analyses).
Confirmatory factor analyses, Asian Indian persons and
White U.K./Irish persons.
Given that the only ethnic groups for
which n‟s exceeded 50 were Indian (n = 88) and White U.K./Irish
persons (n = 54), we estimated a multiple-group confirmatory
factor analysis using LISREL, across these two groups and based
Psychometric Properties
18
on the 3-factor solution obtained from the aforementioned
exploratory factor analysis (see Jackson, 2003, for a recent
commentary regarding sample size and structural equation
analyses).
Correlation matrices for the two groups, based on
the 10 remaining MEIM items, are available from the first author
upon request.
In this multiple-group model, which included
equality constraints on all parameters, Items 1, 2, 3, and 4
were assigned to the Behavioral component of ethnic identity;
Items 6, 8, and 9 were assigned to the Cognitive component of
ethnic identity; Items 7, 10, and 12 were assigned to the
Affective component of ethnic identity; and the three factors
were allowed to correlate.
Unfortunately, the two-group
confirmatory factor analysis produced a fitted covariance matrix
that was not positive definite (i.e., one or more of the
correlations in the fitted covariance matrix were greater than
1.00 and, hence, yielded an inadmissible solution; Schumacker &
Lomax, 2004).
This failure prevented us from performing tests
of either metric invariance (in which factor structures are
identical across groups) or scalar invariance (in which means
are identical across groups; see Cheung & Rensvold, 2002;
Vandenberg & Lance, 2000).
Although the problem regarding the fitted covariance matrix
surfaced when we tried to enter the two correlation matrices in
the same analysis, we did not encounter any problems when we ran
two separate, one-group confirmatory factor analyses.
For Asian
Indian persons, the goodness-of-fit indices clearly supported
Psychometric Properties
19
the 3-factor model, χ2 (32) = 43.61, p < .09; χ2/df = 1.36; RMSEA
= .06; CFI = .99). Inspection of factor loadings (shown in Table
2) indicated that all loadings for items on their associated
factors were greater than .50.
Results of reliability analyses
indicated that the subscales were internally consistent
(Cronbach‟s alphas = .83 for the Behavioral component, .90 for
the Cognitive component, and .92 for the Affective component).
Zero-order correlations among the three components of ethnic
identity were positive and significant (for Behavioral and
Cognitive components, r = .57, p < .01; for Behavioral and
Affective components, r =.53, p < .01; for Cognitive and
Affective components, r = .71, p < .01).
For White U.K./Irish persons, the goodness-of-fit indices
provided mixed support for the 3-factor model, χ2 (32) = 60.18, p
< .01; χ2/df = 1.88; RMSEA = .11; CFI = .95. Inspection of the
factor loadings (Table 2) indicated that all nonzero loadings
were greater than .50.
Results of reliability analyses
indicated that the subscales were internally consistent
(Cronbach‟s alphas = .80 for the Behavioral component, .89 for
the Cognitive component, and .88 for the Affective component).
Zero-order correlations among the three components of ethnic
identity were positive and significant (for Behavioral and
Cognitive components, r = .65, p < .01; for Behavioral and
Affective components, r = .48, p < .01; for Cognitive and
Affective components, r = .75, p < .01).
Discussion
Psychometric Properties
20
Results of the present study indicated that, contrary to the 1factor and 2-factor solutions that commonly have been reported
(Phinney & Ong, 2007), a 3-factor solution provided an optimal
fit to the MEIM data for our U.K. sample as a whole.
The
resulting factors (i.e., Behavioral, Cognitive, and Affective)
bear a strong resemblance to the three dimensions (i.e., Ethnic
Behaviors, Search, and Affirmation) that appeared in the
earliest version of the MEIM (Phinney, 1992; Phinney & Alipuria,
1990).
Although Phinney and Ong (2007) recommended keeping
behavioral manifestations of ethnic identity separate from
cognitive and affective components of ethnic identity, our
results suggest that participants view such behavioral
manifestations as part and parcel of ethnic identity.
Our results regarding Asian Indian persons in particular are at
odds with the results of Berry, Phinney, Sam, and Vedder (2006).
We obtained three intercorrelated factors among Asian Indians in
the U.K., whereas Berry et al. obtained one factor among Asian
Indians in the U.K.
However, Berry et al. used a modified 8-
item version of the MEIM.
As a result, we believe that it would
be premature to draw firm conclusions regarding the differences
between our results and the results obtained by Berry et al.
Our results regarding the applicability of the 3-factor
ethnic identity model to White U.K./Irish persons are novel in
that, to our knowledge, no previously published study has
examined the psychometric properties of the MEIM among this
segment of the U.K. population.
Given that approximately 90% of
Psychometric Properties
21
persons in the U.K identify themselves as White U.K./Irish (see
U.K. Office of National Statistics, 2005), it is important to
assess the structure of ethnic identity in this group.
Of
course, the same problem exists in the U.S.; many people assume
that White Americans do not “have” culture or ethnicity
(Phinney, 1996)4.
Moreover, ethnic minorities make up a much
larger share of the population in the U.S. than in the U.K.
(Kibria, 2007).
Since the late 1990s, right-wing political
groups (especially the British National Party, which ostensibly
has sought mainstream acceptance but often has been viewed as
sympathetic to neo-Nazi propaganda; Copsey, 2007) increasingly
have invoked the need to protect the ethnic identity of the
White U.K./Irish population against a presumed rising tide of
ethnic minority immigrants (Runnymede Trust, 2000). Clearly,
then, the issue of ethnic identity among White U.K./Irish
persons is timely, both within and outside academia.
Strengths and Limitations of the Present Study
The present study is characterized by some important
strengths.
For example, the overall sample (which included
workers as well as students) was highly diverse, in keeping with
the demographic characteristics of West London (e.g., in the
West London borough of Brent, approximately 38% of the
population is classified as European-descent, 44% as Asiandescent, and 18% as African-descent; U.K. Office of National
Statistics, 2006).
In addition, after 1- and 2-factor
confirmatory factor analyses did not work as expected, an
Psychometric Properties
22
exploratory factor analysis yielded a 3-factor solution with
subscales that met the most stringent criteria for reliability
(Carmines & Zeller, 1979) and were highly intercorrelated yet
distinct dimensions.
Finally, to our knowledge, this is the
first study to apply the 12-item MEIM to multiple ethnic groups
in the U.K.
At the same time, the present study is characterized by
some important limitations.
With regard to demographic
characteristics, the ethnic makeup of the present sample does
not match the ethnic makeup of the United Kingdom as a whole (in
which approximately 92% of the population is classified as
European-descent, 4% as Asian-descent, and 2% as African
descent; U.K. Office of National Statistics, 2003).5
Also, with
regard to the 3-factor solution, the fact that previous
researchers have not reported such a solution with the 12-item
MEIM leads one to wonder whether our results generalize beyond
the present sample (although results of an exploratory factor
analysis of the 14-item MEIM in a sample of Asian Americans
yielded the a factor structure with the same three factors that
we identified; Lee & Yoo, 2004).
The most serious limitation,
however, is the lack of metric invariance, which not only
prevented us from making direct comparisons regarding factor
structure across ethnic groups but also prevented us from
obtaining scalar invariance, which in turn prevented us from
testing for mean ethnic group differences on the MEIM items.
Psychometric Properties
23
All in all, we believe that the strengths outweigh the
limitations in the present study.
With regard to demographic
characteristics, although persons of European descent were
underrepresented in the present sample, we were fortunate to
collect data in London, which is home to more than 40% of the
U.K. ethnic minority population (U.K. Office of National
Statistics, 2004).
With regard to the 3-factor solution, the
model clearly could be applied to the sample as a whole.
Finally, with regard to lack of metric invariance, we cannot
make direct comparisons between the factor loadings across Asian
Indians and White U.K./Irish persons; nevertheless, the
Behavioral, Cognitive, and Affective factors all emerged in
analyses for Asian Indians and for White U.K./Irish persons.
Directions for Future Research
Future researchers in the U.K. might wish to evaluate the
psychometric properties of ethnic identity measures other than
the MEIM.
For example, the Ethnic Identity Scale (EIS; Umana-
Taylor, Yazedjian, & Bamaca-Gomez, 2004) was developed as an
alternative to the MEIM, measuring the dimensions of
Exploration, Resolution, and Affirmation.
So far, to our
knowledge, use of the EIS has been limited to the U.S.
As
researchers begin to acknowledge the importance of ethnic
identity in increasingly multiethnic societies such as the U.K.
(Berry et al., 2006), comparisons of the psychometric properties
of various ethnic identity scales will be needed.
Psychometric Properties
24
In addition, future researchers in the U.K. might wish to
evaluate the psychometric properties of scales that were
designed to measure racial identity, rather than other aspects
of ethnic identity (see Worrell & Gardner-Kitt, 2006, regarding
the distinction between racial and ethnic identities).
For
example, the Racial Identity Attitude Scale (RIAS; Parham &
Helms, 1981), designed to measure Pre-Encounter, Encounter,
Immersion, and Internalization stages of Black identity
development, has emerged as the most widely used measure of
racial identity (Helms, 2007).
However, details regarding the
earliest factor analyses of the RIAS in the U.S. were never
published (Cokley, 2007).
Similarly, Robinson (2000)
administered the RIAS to two Black samples in the U.K. but did
not report results of factor analyses.
As researchers begin to
acknowledge the distinctive social and psychological experiences
of European, African, and Asian descent groups in the U.K.
(Modood et al., 1997), critical evaluation of measures of racial
identity will be needed.
Conclusion
At the outset of the present paper, we mentioned that the
1990s were hailed as the “decade of ethnicity” in the U.S.
(Shweder & Sullivan, 1993, p. 517).
Thanks largely to the
efforts of Phinney and her colleagues (e.g., Phinney, 1992;
Phinney & Alipuria, 1990; Phinney, Chavira, & Tate, 1993;
Phinney, Ferguson, & Tate, 1997; Roberts et al., 1999), the
1990s generated a large body of research providing valuable
Psychometric Properties
insights into ethnic identity in the U.S.
25
We hope that the
present findings (as well as the results of Berry et al., 2006)
will help to increase researchers‟ understanding and interest
regarding ethnic identity – whether measured by the MEIM or by
other scales – within the U.K.
Psychometric Properties
26
References
Alexander, C. (2006).
Introduction:
Ethnic and Racial Studies,
Mapping the issues.
29, 397-410.
Avery, D. R., Tonidandel, S., Thomas, K. M., Johnson, C. D., &
Mack, D. A. (2007).
Assessing the Multigroup Ethnic
Identity Measure for measurement
racial and ethnic groups.
equivalence across
Educational and Psychological
Measurement, 67, 877-888.
Baumeister, R. F. (1997).
esteem:
Identity, self-concept, and self-
The self lost and found.
In R. Hogan, J. Johnson,
& S. Briggs (Eds.), Handbook of personality psychology (pp.
681-710).
San Diego:
Academic Press.
Berry, J. W., Phinney, J. S., Sam, D. L., & Vedder, P., eds.
(2006).
Immigrant youth in cultural transition:
Acculturation, identity, and adaptation across national
contexts.
Mahwah, New Jersey:
Erlbaum.
Bhui, K., Lawrence, A., Klineberg, E., Woodley-Jones, D.,
Taylor, S., Stansfeld, S., Viner, R., & Booy, R. (2004).
Acculturation and health status among African-Caribbean,
Bangladeshi and White British adolescents.
Social
Psychiatry and Psychiatric Epidemiology, 40, 259-266.
Bosma, H. A., & Kunnen, E. S. (2001).
Determinants and
mechanisms in ego identity development:
analysis.
A review and
Developmental Review, 21, 39-66.
Psychometric Properties
British Psychological Society (2005).
27
Code of conduct, ethical
principles and guidelines of the British Psychological
Society.
Leicester, United Kingdom:
Brown, T. A. (2006).
research.
BPS.
Confirmatory factor analysis for applied
New York:
Guilford.
Burgess, S., & Wilson, D. (2005).
England‟s schools.
Ethnic segregation in
Transactions of the Institute of
British Geographers, 30, 20-36.
Carmines, E. G., & Zeller, R. A. (1979).
validity assessment.
Cokley, K. (2007).
Newbury Park, California:
Sage.
Critical issues in the measurement of ethnic
and racial identity:
field.
Reliability and
A referendum on the state of the
Journal of Counseling Psychology, 54, 224-234.
Copsey, N. (2007).
Changing course or changing clothes?
Reflections on the
ideological evolution of the British
National Party 1999-2006.
Patterns of
Prejudice, 41, 61-
82.
Cheung, G. W., & Rensvold, R. B. (2002).
Evaluating goodness-
of-fit indexes for testing measurement invariance.
Structural Equation Modeling, 9, 233-255.
Dandy, J., Durkin, K., McEvoy, Barber, B. L., & Houghton, S.
(2008).
Psychometric properties of multigroup ethnic
identity measure (MEIM) scores with Australian adolescents
from diverse ethnocultural groups.
Journal of Adolescence,
31, 323-335.
DeVos, T., & Banaji, 2005).
American = White?
Journal of
Psychometric Properties
28
Personality and Social Psychology, 88, 447-466.
Erikson, E. H. (1950).
Childhood and society.
New York:
Norton.
Erikson, E. H. (1968).
Identity:
Youth and crisis.
New York:
Norton.
Eslea, M., & Mukhtar, K. (2000).
Bullying and racism among
Asian schoolchildren in Britain.
Educational Research, 42,
207-217.
Gaines, S. O., Jr., Marelich, W. D., Bledsoe, K. L., Steers, W.
N., Henderson, M. C., Granrose, C. S., Barajas, L., Hicks,
D., Lyde, M., Takahashi, Y., Yum, N., Rios, D. I., Garcia,
B. F., Farris, K., & Page, M. S. (1997).
Links between
race/ethnicity and cultural values as mediated by
racial/ethnic identity and moderated by gender.
Journal of
Personality and Social Psychology, 72, 1460-1476.
Gilborn, D. (2006).
“WhiteWorld.”
Rethinking White supremacy:
Who counts in
Ethnicities, 6, 318-340.
Goossens, Y. (2006).
Self-esteem and sellf-actualisation as
predictors of accommodation.
Unpublished senior thesis,
Brunel University.
Heer, D. (2006).
Commitment as a mediator of the impact of
self-esteem on accommodation in heterosexual
relationships.
University.
Unpublished senior thesis, Brunel
Psychometric Properties
Helms, J. E. (2007).
29
Some better practices for measuring racial
and ethnic identity constructs.
Journal of Counseling
Psychology, 54, 235-246.
Jackson, D. L. (2003).
Revisiting sample size and number of
parameter estimates:
Some support for the N:q hypothesis.
Structral Equation Modeling, 16, 128-141.
Jöreskog , K., & Sörbom, D. (1996a).
reference guide.
Chicago:
LISREL 8:
User’s
Scientific Software
International.
Jöreskog, K., & Sörbom, D. (1996b).
guide.
Chicago:
PRELIS 2:
User’s reference
Scientific Software International.
Jöreskog, K., & Sörbom, D. (2005a).
LISREL 8.72.
Chicago:
Scientific Software International.
Jöreskog , K., & Sörbom, D. (2005b).
PRELIS 2.72.
Chicago:
Scientific Software International.
Kibria, N. (2007).
The “new Islam” and Bangladeshi youth in
Britain and the US.
Ethnic and Racial Studies, 31, 243-
266.
Kimmel, A. J. (2004).
research.
Ethical issues in social psychology
In C. Sansone, C. C. Morf, & A. T. Panter
(Eds.), The SAGE handbook of methods in social psychology
(pp. 45-70).
Kline, P. (1994).
Thousand Oaks, California:
SAGE.
An easy guide to factor analysis.
London:
Routledge.
Kline, R. B. (2005).
Principles and practices of structural
equation modelling (2nd ed.).
New York:
Guilford.
Psychometric Properties
Large, P., & Gnosh, K. (2006).
30
Estimates of the polutaion by
ethnic group for areas within England.
Population Trends,
124, 8-17.
Lee, R. M., & Yoo, H. C. (2004).
Structure and measurement of
ethnic identity for Asian American college students.
Journal of Counseling Psychology, 51, 263-269.
Lidder, A. (2006).
Commitment, gender, and ethnicity as
predictors of accommodation.
Unpublished senior thesis,
Brunel University.
Luyckx, K., Goossens, L., Soenens, B., & Beyers, W. (2006).
Unpacking commitment and exploration:
Preliminary
validation of an integrative model of late adolescent
identity formation.
Mann, A. (2006).
Journal of Adolescence, 29, 361-378.
Ethnicity as a predictor of accommodation in
adult relationships with cultural values as covariates.
Unpublished senior thesis, Brunel University.
Marcia, J. E. (1966). Development and validation of ego-identity
status.
Journal of Personality and Social Psychology, 3,
551-558.
Marcia, J. E. (1967).
Ego identity status:
Relationship to
change in self-esteem, „general maladjustment,‟ and
authoritarianism.
Marcia, J.E. (1980).
Journal of Personality, 35, 119-133.
Identity in adolescence.
In J. Andelson
(Ed.), Handbook of adolescent psychology (pp. 159-187).
New York:
Wiley.
Marcia, J. E., & Friedman, M. L. (1970).
Ego identity status in
Psychometric Properties
college women.
31
Journal of Personality, 38, 249-263.
Marsh, H. W., Balla, J. R., & McDonald, R. P. (1988).
of-fit indices in confirmatory factor analysis:
of sample size.
Minhas, S. (2006).
GoodnessThe effect
Psychological Bulletin, 103, 391-410.
Ethnic group membership as a predictor of
components of ethnic identity in the United Kingdom.
Unpublished senior thesis, Brunel University.
Modood, T., Berthoud, R., Lakey, J., Nazroo, J., Smith, P.,
Virdee, S., & Beishon, S. (1997).
Britain:
Ethnic minorities in
Diversity and disadvantage.
London:
Policy
Studies Institute.
Nagel, J. (2003).
Race, ethnicity, and sexuality:
intersections, forbidden frontiers.
New York:
Intimate
Oxford
University Press.
Orlofsky, J. L., Marcia, J. E., & Lesser, I. M. (1973).
Ego
identity status and the intimacy versus isolation crisis of
young adulthood.
Journal of Personality and Social
Psychology, 27, 211-219.
Parham, T. A., & Helms, J. E. (1981).
The influence of Black
students‟ racial identity attitudes on preferences for
counselor‟s race.
Journal of Counseling Psychology, 28,
250-257.
Peach, C. (2006).
Islam, ethnicity and South Asian religions in
the London 2001 census.
Transactions of the Institute of
British Geographers, 31, 353-370.
Psychometric Properties
Pegg, P. O., & Plybon, L. E. (2005).
measurement of ethnic identity.
32
Toward the theoretical
Journal of Early
Adolescence, 25, 250-264.
Phinney, J. S. (1990).
adults:
Ethnic identity in adolescents and
A review of research.
Psychological Bulletin,
108, 499-514.
Phinney, J. S. (1992).
The Multi-group Ethnic Identity Measure:
A new scale for use with adolescents and you adults from
diverse groups. Journal of Adolescent Research, 7, 156-176.
Phinney, J. S. (1996).
When we talk about American ethnic
groups, what do we mean?
American Psychologist, 51, 918-
927.
Phinney, J. S., & Alipuria, L. L. (1990).
college students.
Ethnic identity in
Journal of Adolescence, 13, 171-183.
Phinney, J. S., & Alipuria, L. L. (1996).
At the interface of
cultures:
Multiethnic/multiracial high school and college
students.
Journal of Social Psychology, 136, 139-158.
Phinney, J. S., Chavira, V., & Tate, J. D. (1993).
The effect
of ethnic threat on ethnic self-concept and own-group
ratings.
Journal of Social Psychology, 133, 469- 478.
Phinney, J. S., Ferguson, D. L., & Tate, J. D. (1997).
Intergroup attitudes among ethnic minority adolescents:
causal model.
Child Development, 68, 955-969.
Phinney, J. S., & Ong, A. D. (2007).
measurement of ethnic identity:
directions.
A
Conceptualization and
Current status and future
Journal of Counseling Psychology, 54, 271-281.
Psychometric Properties
33
Ponterotto, J. G., Gretchen, D., Utsey, S. O., Stracuzzi, T., &
Saya, R., Jr. (2003).
Measure (MEIM):
testing.
The Multigroup Ethnic Identity
Psychometric review and further validity
Educational and Psychological Measurement, 63,
502-515.
Reese, L. E., Vera, E. M., & Paikoff, R. L. (1998).
identity assessment among inner-city children:
Ethnic
Evaluating
the applicability of the Multigroup Ethnic Identity
Measure.
Journal of Black Psychology, 24, 289-304.
Roberts, R., Phinney, J., Masse, L., Chen, Y., Roberts, C., &
Romero, A. (1999).
The structure of ethnic identity in
young adolescents from diverse ethnocultural groups.
Journal of Early Adolescence, 19, 301-322.
Robinson, L. (2000).
Racial identity attitudes and self-esteem
of Black adolescents in residential care:
study.
British Journal of Social Work, 30, 3-24.
Runnymede Trust (2000).
London:
An exploratory
The future of multi-ethnic Britain.
Profile Books.
Schenkel, S., & Marcia, J. E. (1972).
Attitudes toward
premarital intercourse in determining ego identity status
in college women.
Journal of Personality, 40, 472-482.
Schumacker, R. E., & Lomax, R. G. (2004).
A beginner’s guide to
structural equation modeling (2nd ed.). Mahwah, New Jersey:
Erlbaum.
Psychometric Properties
Schwartz, S. J. (2001a).
Author‟s response:
identity: A rejoinder.
Identity:
34
The evolution of
An International Journal
of Theory and Research, 1, 87-93.
Schwartz, S. J. (2001b).
The evolution of Eriksonian and neo-
Eriksonian identity theory and research:
integration.
Identity:
A review and
An International Journal of Theory
and Research, 1, 7-58.
Schwartz, S. J. (2007).
consolidation:
The structure of identity
Multiple correlated constructs or one
superordinate construct?
Identity, 7, 27-49.
Shweder, R. A., & Sullivan, M. A. (1993).
Who needs it?
Cultural psychology:
Annual Review of Psychology, 44, 497-523.
Spencer, M. S., Icard, L. D., Harachi, T. W., Catalano, R. F., &
Oxford, M. (2000).
Ethnic identity among monoracial and
multiracial early adolescents.
Journal of Early
Adolescence, 20, 365-387.
Thompson, B. (2004).
analysis.
Exploratory and confirmatory factor
Washington, DC:
American Psychological
Association.
Toder, N. L., & Marcia, J. E. (1973).
Ego identity status and
response to conformity pressure in college women.
of Personality and Social Psychology, 26,
287-294.
Tsai, J. L., Mortensen, H., Wong, Y., & Hess, D. (2002).
does “being American” mean?
Journal
What
A comparison of Asian American
and European American young adults.
Cultural Diversity and
Ethnic Minority Psychology, 8, 257-273.
Psychometric Properties
35
U.K. Office of National Statistics (2003, February 13).
Population size:
7.9% from a minority ethnic group.
Retrieved February 26, 2008, from
http://www.statistics.gov.uk/cci/nugget_print.asp?ID=273
U.K. Office of National Statistics (2004, January 8).
Geographic distribution:
London.
45% of non-White people live in
Retrieved February 26, 2008, from
http://www.statistics.gov.uk/cci/nugget_print.asp?ID=457
U. K. Office of National Statistics (2005, December 15).
Foreign-born:
1 in 12 in UK born overseas.
Retrieved
September 2, 2008, from
http://www.statistics.gov.uk/cci/nugget_print.asp?ID=1312
U.K. Office of National Statistics (2006, December 15).
Geographic diversity:
area.
Brent the most ethnically diverse
Retrieved February 26, 2008, from
http://www.statistics.gov.uk/CCI/nugget_print.asp?ID=1700
Umana-Taylor, A. J., Yazedijan, A., & Bamaca-Gomez, M. (2004).
Developing the Ethnic Identity Scale:
social identity perspectives.
Using Eriksonian and
Identity:
An International
Journal of Theory and Research, 4, 9-38.
Vandenberg, R. J., & Lance, C. E. (2000).
A review and
synthesis of the measurement invariance literature:
Suggestions, practices, and
organizational research.
3, 4-70.
recommendations for
Organizational Research Methods,
Psychometric Properties
Verkuyten, M. (2005).
Hove:
36
The social psychology of ethnic identity.
Psychology Press.
Verkuyten, M., & Yildiz, A. A. (2006).
minority rights:
The endorsement of
The role of group position, national
context, and ideological beliefs.
Political Psychology,
27, 527-659.
Worrell, F. C., & Gardner-Kitt, D. L. (2006).
The relationship
between racial and ethnic identity in Black adolescents:
The Cross Racial Identity Scale and the Multigroup Ethnic
Identity Scale.
Identity: An International Journal of
Theory and Research, 6, 293-315.
Yancey, A. K., Aneshensel, C. S., & Driscoll, A. K. (2001).
assessment of ethnic identity in a diverse urban youth
population.
Journal of Black Psychology, 27, 190-208.
The
Psychometric Properties
37
Authors‟ Note
The authors thank Editor Alan Waterman, Action Editor Seth
Schwartz, and three anonymous reviewers for their constructive
comments on earlier versions of this paper.
Psychometric Properties
38
Footnotes
1
The ethnic classification system used in the present study,
including the specific wording, is identical to the ethnic
classification system that the government of the United Kingdom
used in the 2001 census (U.K. Office of National Statistics,
2003).
2
As Marsh, Balla, and McDonald (1988) pointed out, chi-squares
accompanying poorly fitting models are sensitive to sample size,
such that (a) misleadingly low chi-squares can be generated by
small data sets (n‟s below 100); and (b) misleadingly high chisquares can be generated by large data sets (n‟s above 200).
3
CFI and other incremental fit indices were not produced by
PRELIS and, hence, are not reported here.
4
The authors are indebted to Seth Schwartz for noting the
similarity in arguments between White Britons‟ and White
Americans‟ ethnicity.
5
From 2001 to 2003, the percentage of the U.K.‟s “non-White
British” population living in London has declined from 44.7% to
42.5% (Large & Gnosh, 2006).
As Action Editor Seth Schwartz
observed, “I wonder… how much the demographics of London reflect
what other parts of the UK will look like in 10-20 years.
Traditional migration patterns are such that the descendants of
immigrants often move out of the cities and into other areas.”
Psychometric Properties
Table 1:
Matrix of Factor Loadings for MEIM Items, All Participants (n = 234)
Factor
Item
1 (Behavioral)
2 (Cognitive)
3(Affective)
1
.68
.05
.04
2
.61
-.02
.01
3
.65
.15
-.13
4
.95
-.18
.03
5
.35
.11
.41
6
.28
.48
.13
7
-.02
.15
.69
8
-.10
.98
.05
9
.09
.64
.23
10
.10
-.07
.87
11
.19
.38
.38
12
-.04
.01
.91
39
Psychometric Properties
40
NOTE: Within each row, the factor loading with absolute value of .30 or higher is in
boldface (maximum of one loading in boldface per row).
1 = I have spent time trying to find out more about my ethnic group, such as its history,
traditions, and customs. (originally Exploration item 1)
2 = I am active in organizations or social groups that include mostly members of my own
ethnic group. (originally Exploration item 2)
3 = I think a lot about how my life will be affected by my ethnic group membership.
(originally Exploration item 3)
4 = In order to learn more about my ethnic background, I have often talked to other people
about my ethnic group. (originally Exploration item 4)
5 = I participate in cultural practices of my own group, such as special food, music, or
customs. (originally Exploration item 5)
6 = I have a clear sense of my ethnic background and what it means for me. (originally
Commitment item 1)
7 = I have a strong sense of belonging to my own ethnic group. (originally Commitment
item 3)
8 = I understand pretty well what my ethnic group membership means to me. (originally
Commitment item 4)
9 = I am happy that I am a member of the group I belong to. (originally Commitment item
2)
10 = I have a lot of pride in my ethnic group. (originally Commitment item 5)
11 = I feel a strong attachment towards my own ethnic group. (originally Commitment item
6)
12 = I feel good about my cultural or ethnic background. (originally Commitment item 7)
Psychometric Properties
Table 2:
Matrix of Factor Loadings for MEIM Items
-------------------------------------------------------------------------------------------------------Asian Indian persons (n = 88)
-------------------------------------------------------------------------------------------------------Factor
Item
1 (Behavioral)
1
.82
2
.75
3
.57
4
.86
2 (Cognitive)
3(Affective)
5
6
.90
7
.78
8
.92
9
.88
10
.91
11
12
.87
41
Psychometric Properties
42
NOTE: Within each row, the factor loading with absolute value of .30 or higher is in
boldface (maximum of one loading in boldface per row).
1 = I have spent time trying to find out more about my ethnic group, such as its history,
traditions, and customs. (originally Exploration item 1)
2 = I am active in organizations or social groups that include mostly members of my own
ethnic group. (originally Exploration item 2)
3 = I think a lot about how my life will be affected by my ethnic group membership.
(originally Exploration item 3)
4 = In order to learn more about my ethnic background, I have often talked to other people
about my ethnic group. (originally Exploration item 4)
5 = I participate in cultural practices of my own group, such as special food, music, or
customs. (originally Exploration item 5)
6 = I have a clear sense of my ethnic background and what it means for me. (originally
Commitment item 1)
7 = I have a strong sense of belonging to my own ethnic group. (originally Commitment
item 3)
8 = I understand pretty well what my ethnic group membership means to me. (originally
Commitment item 4)
9 = I am happy that I am a member of the group I belong to. (originally Commitment item
2)
10 = I have a lot of pride in my ethnic group. (originally Commitment item 5)
11 = I feel a strong attachment towards my own ethnic group. (originally Commitment item
6)
12 = I feel good about my cultural or ethnic background. (originally Commitment item 7)
Psychometric Properties
-------------------------------------------------------------------------------------------------------White UK/Irish persons (n = 54)
-------------------------------------------------------------------------------------------------------Factor
Item
1 (Behavioral)
1
.67
2
.54
3
.69
4
.99
2 (Cognitive)
3(Affective)
5
6
.78
7
.93
8
.90
9
.78
10
.87
11
12
.89
43
Psychometric Properties
44
NOTE: Within each row, the factor loading with absolute value of .30 or higher is in
boldface (maximum of one loading in boldface per row).
1 = I have spent time trying to find out more about my ethnic group, such as its history,
traditions, and customs. (originally Exploration item 1)
2 = I am active in organizations or social groups that include mostly members of my own
ethnic group. (originally Exploration item 2)
3 = I think a lot about how my life will be affected by my ethnic group membership.
(originally Exploration item 3)
4 = In order to learn more about my ethnic background, I have often talked to other people
about my ethnic group. (originally Exploration item 4)
5 = I participate in cultural practices of my own group, such as special food, music, or
customs. (originally Exploration item 5)
6 = I have a clear sense of my ethnic background and what it means for me. (originally
Commitment item 1)
7 = I have a strong sense of belonging to my own ethnic group. (originally Commitment
item 3)
8 = I understand pretty well what my ethnic group membership means to me. (originally
Commitment item 4)
9 = I am happy that I am a member of the group I belong to. (originally Commitment item
2)
10 = I have a lot of pride in my ethnic group. (originally Commitment item 5)
11 = I feel a strong attachment towards my own ethnic group. (originally Commitment item
6)
12 = I feel good about my cultural or ethnic background. (originally Commitment item 7)